|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
154.20.79.52
Hi folks,
I posted a thread about 10 days ago about sibilance and plate voltage. Got a couple of nice suggestions from two fellow inmates. I am thinking that I have the filament and plate voltages about where they should be now. To recap....I have a wonderful custom push pull amp. Cathode bias, no feedback, AC filament. I generally use EL34's, but it also has a switch position to change the bias setting for 6550/KT88's. After enjoying it for about 3 years, the power transformer decided to die. The designer dug up another old stock transformer for me and did the labor for free. This transformer is huge - probably almost twice as large as the original - also old stock. Anyway, I just can't get the sound of the amp back the way I had it. It is a bit bright, and sibilance has reared its ugly head! I never had even a hint of sibilance previously. Nothing else in the system has changed. So here is what I have done so far:
Rolled all sorts of tubes from the stash with predictable changes to the tonal balance, but overall the amp is still a bit bright (even with OS Tungsol 6550’s) and has some sibilance I cannot get rid of.
Filament voltage is now 6.1 volts and B+ is 423 volts. I think they are fine.
I have not played at all with the coupling caps. Because of their mounting, and my general inexperience I cannot identify the brand and type. The second stage coupling caps (I think) are the larger blue ones in the picture (upper left). Writing is faint, but I think Philips is the brand. 275 volts, and I think it says 470n so my guess is that is 0.47 micro farad? I can get the specs from the amps designer. The small caps for the first stage – 12au7 – (bottom right) are mounted upside down so I can’t get the brand or value. They look like they may also be made by same manufacturer of the larger ones (Philips?). The brown ones are small Russian NOS PIO’s to sweeten the sound. They read 0.01mK which I assume are .01 mcro farad? No matter, not going to change them. My questions are these.
1. Anyone know what the small first stage caps are (blue, rectangular, bottom right)? Another Philips? What type are these caps – clearly not PIO…? Polyprop? Again, I can get the value from the amp’s designer. Can anyone speculate as to whether changing them will improve the sibilance problem? Are they lousy caps, or mid grade? Amp was wonderful before, but new power X-former although more powerful, has made it brighter. Would be happy to solder some new caps in, but would like recommendations. I don’t mind spending $10 - $20 a pop for the right caps. I need to try to get rid of sibilance, and maybe darken things VERY slightly. However, amp has tons of air and I don’t want to lose that. The speed is about medium and I don’t want to slow it down.
2. Or are the caps probably not going to tame the sibilance….or improve the amp? Barking up the wrong tree?
Any recommedations appreciated. Transformer is quiet, amp is dead quiet, tons of headroom, tons of air, huge soundstage (floor to ceiling and a foot or two wider than speakers, just now some sibilance... Sorry for the picture quality – no macro setting on my camera.
Follow Ups:
Hi.With a larger transformer the internal impedences will be lower, this gives less damping to the resonance between the leakage inductance and winding capacitance, this might be ringing.
Try adding a 100R resistor in series with 0.1uF 1000V capacitor across the HT winding (if the winding has a CT then add to each half.)
Hope this helps.
Chris
Chris,
Added the cap and resistor to each side. No noticeable effect on the sibilance, but it made things a bit more focused. Cleaned things up a bit. Must have been a bit of ringing or some sort of grunge in the power supply. A nice worthwhile tweak anyway for a few dollars worth of parts and 10 minutes time! Guess I will lower the B+ by 10 volts or so as Jimmy suggested by using a 50 ohm, 20 watt resistor on each side. Right now it reads 423 volts of B+. Don't want to go too much lower as I am running in triode... but it will be a cheap experiment that I can undo if I don't like it or it has no effect. I will leave your little filter in there though!Thanks for the tip.
Hi Chris,
Thanks. Will pay a call to the amp's designer with notes in hand. We can probably wire that up in a jiffy.....will let you know.
Hi Don,I got to say this, and please accept it as constructive criticism – you’re shooting in the dark! You changed the power transformer, which changed the sound, and now you are trying all sorts of things to fix it. Your problem has to be associated with the difference between the two transformers. For example, perhaps the resistance of the high voltage winding is less in the new transformer than in the old one. I would start by making a list of all of the differences and then compensate for those differences one-by-one.
Another alternative would be to have your old transformer repaired. There are places in most metropolitan areas that will rewind motors and transformers. Look in your Yellow Pages.
Hi Dave,
No offence taken! You are right, that is precisely what I am trying to do. I realize that the only difference is the transformer. I know exactly how I am trying to get it to sound. I have had stereos in one incarnation or another for 38 years and never had a sibilance problem. Now I have one. There is some sort of peak in the upper midrange now where the sibilance lives. I have tried darkening things a bit with my stash of tubes but all I can do is make predictable changes in the tonal balance. The warm or dark tubes sound warm or darker, the bright tubes brighter, but no matter what there is sibilance. I have made sure the voltages are where they should be or were before. I have tried various filament voltages as an experiment, but can't tame the sibilance. Transformers are rather expensive so I would rather not change it out, but it looks like that is where I am headed. Rewinding the old one is not a bad suggestion, although it may have died because it could have been slightly undersized, who knows. Or maybe it's number was just up.... The current one doesn't even get warm. I was just thinking that a different coupling cap might do the trick for less money... You are right, the difference is the transformer and I am trying to compensate...
Hi Don,Here is another suggestion. Assuming that there is a peak in the upper midrange perhaps there is a resonance in the power supply.
Voltsecond has a very good article there about resonances in the power supply.http://www.siteswithstyle.com/VoltSecond/CLC_ringing/CLC_ringing.html
Dave,
Thanks for the link. Had a quick look. Very interesting article. Will have to discuss it with the amp's designer. Darn thing sounds great as long as you don't listen to vocals. Instrumentals are great. Just a hint of sibilance on vocals to drive you nuts. On the well recorded vocals there is little if any, but most pop recordings...By the way, the Philips caps are polypropylene....
With all good intentions, & meant to be constructive: in the pic alone there is a resistor blob-soldered to a metal plate, melted wire insulation from a solder iron getting too close, a circut Common Buss wrigling wily-nily thru parts, and a general mess of wiring. This *custom designer* has poor assembly skills IMO & is thus more likely a newbee hobbiest rather than a qualified EE. Re-consider what he did with the mis-matched pair OPT replacement (your present problem) before visiting him again for additional errors. With notes or not, that so-called *custom designer* guy does not know what he is doing.
Did I miss something? What was the orginal B+ voltage?
Hi Jimmy,
Well, I tried a 50 ohm and then a 100 ohm resistor. B+ is now 420 by my meter and was about 433 ahead of the resistor. So a drop of 13+ volts and I can hear no change in the sibilance....probably not related to B+....Must just be something in the sound of that transformer..will keep working on it. The power supply filter didn't affect sibilance either.
This post is made possible by the generous support of people like you and our sponsors: